Locating hyperfunctioning parathyroid glands using 11C-Choline PET/CT: an inter- and intra-observer variation study

Julie Wulf Christensen; Lars Thorbjørn Jensen; Susanne Bonnichsen Søndergaard; Rikke Broholm; Christian Haarmark; Martin Krakauer; Finn Noe Bennedbæk; Bo Zerahn; Waldemar Trolle; Christoffer Holst Hahn; Bent Kristensen

doi:10.1186/s41824-021-00108-z

Locating hyperfunctioning parathyroid glands using ¹¹C-Choline PET/CT: an inter- and intra-observer variation study

Eur J Hybrid Imaging. 2021 Jul 6;5(1):13. doi: 10.1186/s41824-021-00108-z.

Affiliations

¹ Department of Nuclear Medicine, Herlev and Gentofte Hospital, 2730, Herlev, Denmark. juliewulfc@gmail.com.
² Department of Nuclear Medicine, Herlev and Gentofte Hospital, 2730, Herlev, Denmark.
³ Department of Clinical Physiology and Nuclear Medicine, Bispebjerg and Frederiksberg Hospital, 2400, Copenhagen, Denmark.
⁴ Department of Medicine, Division of Endocrinology, Herlev and Gentofte Hospital, 2730, Herlev, Denmark.
⁵ Department of Otorhinolaryngology and Neck Surgery, North Zealand Hospital, 3400, Hilleroed, Denmark.
⁶ Department of Otorhinolaryngology and Neck Surgery, Rigshospitalet, 2100, Copenhagen, Denmark.

Abstract

Background: Use of ¹¹C-Choline PET/CT is gaining ground in detecting hyperfunctioning parathyroid glands in primary hyperparathyroidism. The purpose of this study was to evaluate the robustness of ¹¹C-Choline PET/CT by assessing intra- and inter-observer agreement to determine whether the method was reader sensitive and therefore should only be performed at highly specialised sites with a high number of cases. PET/CT images of 40 patients diagnosed with primary hyperparathyroidism were anonymised and evaluated three times by three readers: an expert reader and two non-experts (non-experts were experienced in PET/CT imaging, but not in ¹¹C-Choline PET/CT in the setting of primary hyperparathyroidism). Number of hyperfunctioning parathyroid glands, location relative to the thyroid gland and confidence of each assessment (low, moderate or high) were noted, and intra- and inter-observer agreement calculated using Fleiss' kappa method. Sensitivities and specificities of the non-experts were calculated using the expert reader as gold standard.

Results: Intra-observer agreement was 'good' to 'near perfect' for all readers. Inter-observer agreement was good between non-experts and the expert, with kappa values ≥ 0.74. Sensitivities between non-experts and the expert were high, > 81%, when assessing which side and 75% when assessing thyroid quadrant. All specificities were > 94%. Reader certainties were 'high' in > 80% of cases for the expert and > 70% and > 65%, respectively for the two non-experts.

Conclusion: ¹¹C-Choline PET/CT is not reader sensitive for the localisation of hyperfunctioning parathyroid glands and may therefore be safely implemented at sites that have a moderate number of cases. Access to a cyclotron laboratory is, however, a necessity for the production of ¹¹C-Choline. The study was conducted in accordance with the Helsinki 2 declaration and The International Council for Harmonisation Guideline for Good Clinical Practice (ICH_GCP) clinical trial, approved by the Research Ethics Committee of the Capital Region of Denmark (Journal-nr.:H-18012490, date of approval: 18 June 2018) and the Danish Medicines Agency (EudraCT no. 2018-000726-63, date of approval: 6 June 2018). The GCP unit in Eastern Denmark has carried out regular monitoring of the trial according to GCP (ID: 2018-1050).

Keywords: 11C-Choline PET/CT; Hyperparathyroidism; Inter-observer agreement; Intra-observer agreement.